You can deploy a custom private model using one of the available open-source models or an integrated API model. You will need a Tune Studio Pro account to deploy an open-source model like Llama 3 or Open Hermes 2, or an API key from the provider to deploy an integrated API model like Claude 3 Opus or ChatGPT-4.

To deploy an integrated API model, navigate to the Models tab in the Tune Studio dashboard.

Click the + Deploy model button to open the deploy model dialog.

Select the model provider from the options in the Get Started section. Enter your API key and click Create. In this example, we selected OpenAI.

Select the model to use (for example, GPT-4), give your custom model a name, and click Add Model.

Your new custom model will be created and brought online.

Use the custom model’s Model ID in text and chat completions to generate logs. Locate the Model ID by clicking the three-dot icon next to the model name and selecting Dashboard from the dropdown.

The Model ID can be copied from the model detail view.